SD have been top in 5/50 polls in 2018, all conducted by YouGov and Sentio. https://t.co/WnRB9dk9Vg
— Chris Hanretty is staying home (@chrishanretty) July 26, 2018
Sweden has a general election on 9 September 2018. You might have picked up on it because the newspapers have been drawing attention to the prospects of the Sweden Democrats as they are apparently rising in the polls. For those that aren’t familiar with them, the Sweden Democrats are the descendant of the fascist movement but now looks like a fairly standard nationalist anti-immigration party. Could such a party top the polls in a country famed for its social democratic inclusiveness?
Stop me if you think you’ve heard this story before. Geert Wilders’ party was poised to become the largest party in the Dutch elections (it didn’t). Marine Le Pen was poised to top the French presidential poll in the first round (she didn’t). The undoubted improvements by the hard right are regularly over-egged in advance.
This has proved profitable for cool-headed bettors in the past. I wrote about this in advance of the Dutch elections. Betting against Marine Le Pen also proved a winning strategy. In both cases the odds on the hard right were far too short.
Are we looking at the same story this time? Perhaps. Unlike the Dutch elections, there are recent polls showing the Sweden Democrats in the lead. The Chris Hanretty tweet at the top of the thread sums up the position. So the Sweden Democrats are in with a shout of doing this. Favouring the Sweden Democrats in this contest at this stage would mean favouring the methods of two of the regular pollsters in opposition to the results of the majority of pollsters.
Would that be justified? As always, we need to look at the odds. As I write, you can back Sweden Democrats to finish with most seats at 1.83 on Betfair and lay them at 1.84. That is eyebrow-raising, given that they have only been ahead in four out of the last 50 polls (and tied in a fifth). You’d have to have strong views about a quality gap in the methods of polling companies to justify that price. There’s no particular reason to assume that the Sweden Democrats’ support is being systematically underrecorded; that wasn’t the experience in the Netherlands or France.
Are there any indications as to which pollsters to back? Picking winners is a mug’s game (polls are snapshots not predictions and demographic changes may alter the reliability of polls in ways which are impossible to predict in advance). Nevertheless, we have a small clue.
In the wake of the last general election in 2014, all the current pollsters took their usual polls. Sentio and YouGov immediately found an upsurge in support for the Sweden Democrats from the general election that none of the other pollsters picked up at that time. That perhaps suggests that Sentio and YouGov might be overstating Sweden Democrat support.
This isn’t anything like a sure thing. But at 1.84 (or indeed at anything close to evens) the Sweden Democrats look like a clear lay for most seats. At least till new information makes me rethink, I’m betting accordingly.